Re: objtool warning for next-20221118

[Josh Poimboeuf]

On Wed, Nov 23, 2022 at 11:12:42AM -0800, Paul E. McKenney wrote:
> On Wed, Nov 23, 2022 at 10:19:41AM -0800, Josh Poimboeuf wrote:
> > On Wed, Nov 23, 2022 at 09:49:51AM -0800, Paul E. McKenney wrote:
> > > > > > Perhaps the best way would be to stick a REACHABLE annotation in
> > > > > > arch_cpu_idle_dead() or something?
> > > > > 
> > > > > When I apply this on -next, I still get the objtool complaint.
> > > > > Is there something else I should also be doing?
> > > > 
> > > > Silly GCC is folding the inline asm.  This works (but still doesn't seem
> > > > like the right approach):
> > > > 
> > > > diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
> > > > index 26e8f57c75ad..128e7d78fedf 100644
> > > > --- a/arch/x86/kernel/process.c
> > > > +++ b/arch/x86/kernel/process.c
> > > > @@ -702,7 +702,7 @@ static void (*x86_idle)(void);
> > > >  #ifndef CONFIG_SMP
> > > >  static inline void play_dead(void)
> > > >  {
> > > > -	BUG();
> > > > +	_BUG_FLAGS(ASM_UD2, 0, ASM_REACHABLE);
> > > >  }
> > > >  #endif
> > > 
> > > I tried this, and still get:
> > > 
> > > vmlinux.o: warning: objtool: do_idle+0x156: unreachable instruction
> > > 
> > > Maybe my gcc is haunted?
> > 
> > Weird, it worked for me.  I have
> > 
> >   gcc version 12.2.1 20220819 (Red Hat 12.2.1-2) (GCC)
> 
> Me, I have these, so quite a bit older:
> 
> gcc version 8.5.0 20210514 (Red Hat 8.5.0-15) (GCC)
> gcc version 9.4.0 (Ubuntu 9.4.0-1ubuntu1~20.04.1)
> 
> > and I can't really fathom why that wouldn't work.  Maybe it's a
> > different issue?  The "unreachable instruction" warning is limited to
> > one, so when a first warning gets fixed, a second warning might suddenly
> > become visible.
> > 
> > Can you attach arch/x86/kernel/process.o?
> 
> Attached!

Hm, for whatever reason, that .o file is indeed missing the reachable
annotation.  <scratches head>

I confirmed the patch also fixes the warning with:

  gcc version 8.5.0 20210514 (Red Hat 8.5.0-10) (GCC)

No idea why it's not working for you... but maybe it doesn't matter as
I'm still thinking we should go with a different approach.

-- 
Josh